The enormous complexity of a semiconductor manufacturing system is the main obstacle for making quality manufacturing control decisions. Conventional methodologies, such as queueing network and simulation analysis, are often too complex to be used effectively. In this paper, we will demonstrate a methodology to build simple statistical models that faithfully characterize the manufacturing system. We then show how these models can help improve the quality of manufacturing control decisions
Published in:
Semiconductor Manufacturing, 2000. Proceedings of ISSM 2000. The Ninth International Symposium on
Date of Conference: 2000